Income stagnation and housing affordability in the United States

Abstract

Between 1980 and 2016 the share of households in the bottom quintile of the income distribution that owned their own home declined by 10 percentage points. For the same households, the share of monthly income spent on rent increased from 28% in 1960 to over 42% in 2016. To asses the extent to which income stagnation is responsible for the decline in affordability, I use Census microdata to construct counterfactual simulations that capture the evolution of housing market trends under alternative assumptions about the distribution of income. Income stagnation explains nearly the entire decline in affordability for the bottom quintile. Housing market frictions that cause the price of housing to deviate from marginal cost matter more to households further up the income distribution. Using Atkinson-type welfare-based inequality measures, I find that the counterfactual distribution of income – with inequality held constant – results in greater welfare for nearly all possible levels of inequality aversion.

Keywords:

• Housing affordability
• Inequality
• Wage stagnation
• Homeownership

JEL CLASSIFICATIONS:

• R31
• D63
• D31

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Include a link to supplemental dataset at Mendeley Data: http://dx.doi.org/10.17632/n8hhv2ndmd.2

Notes

1 There is also a macroeconomic literature focusing on the opposite channel: the link from housing to income and wealth inequality. Appendix 3 offers a brief discussion of the implications of this paper for that literature.

2 For series available prior to 1950, trends in the variables of interest are not substantially changed if the occupational earnings score is used to sort households into income quintiles in the pre-1950 period and total family income is used to sort households in the post-1950 period. See Figure A1. In the main text, the measure used to sort households into quintiles is kept constant within each figure unless otherwise noted.

3 The original FHA mortgage was a 20-year, fully amortizing loan with a maximum loan-to-value (LTV) ratio of 80% – a duration 10–15 years longer than the average mortgage duration for commercial bank mortgages during the 1920–1939 period, and an LTV ratio 20–30% points larger than the average loan-to-value ratio (Chambers et al., 2014). The Servicemen's Readjustment Act of 1944 (the ‘GI Bill’) provided additional incentives for homeownership by offering no-downpayment-required mortgages to returning World War II veterans. Fetter (2010) provides evidence that the World War II and Korean War GI Bills explain 7.4% of the increase in the homeownership rate over the 1940–1960 period. The joint impact of these changes is evident in both the increase in the rate of homeownership and the shift in the way homeownership is financed in the post-war period.

4 Tax reform in the 1980s – notably the 1981 Economic Recovery Tax Act and the Tax Reform Act of 1986 – reduced tax incentives for rental housing construction (see Figure 2(e)), and diminished the tax-favored status of owner-occupied housing relative to other investments (Poterba, 1994). Actions taken by the Federal National Mortgage Association and the Federal Home Loan Mortgage Corporation encouraged the expansion of secondary mortgage markets which facilitated trading in mortgage-backed securities. Finally, deregulation of financial markets, including the removal of interest rate ceilings on banks competing with savings and loan institutions reduced the importance of these housing-oriented institutions and subsumed housing finance within financial markets more broadly. The securitization and deregulation of mortgage finance can be understood as effects of a broader process of financialization – the growing influence of financial actors and institutions on the rest of the economy (Davis, 2017; Mason et al., 2018) – as it plays out in the market for housing.

5 This measure adjusts for changes in household size by dividing income by the square root of the number of people in the household.

6 https://www.census.gov/construction/cpi/.

7 E.g. agglomeration externalities in coastal cities may contribute to rising income inequality and declining housing affordability, but the methodology applied here does not address underlying causal mechanisms as such.

8 A further caveat is that the impact of the removal of restrictions on housing supply will depend on whether new supply takes the form of multi- or single-family units. It is likely that the benefits to households at the bottom of the income distribution will be larger in the former case.

9 It becomes difficult to make welfare comparisons using $I_{\mathrm{A}\mathrm{t}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{o}\mathrm{n}}$ across distributions involving different values for μ, as a given distribution of income might be characterized by both greater inequality (larger $I_{\mathrm{A}\mathrm{t}\mathrm{k}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{o}\mathrm{n}}$) and greater welfare (larger θ).

10 California Senate Bill 50 (2018). https://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201920200SB50.

11 Income net of housing expenditure.

Additional information

Notes on contributors

Luke Petach

Luke Petach, PhD is an Assistant Professor of Economics in the Jack C. Massey College of Business at Belmont University. His research interests include long-run economic growth, income inequality, and regional economic development.